250 THE GENERAL THEORY OF EMPLOYMENT bk. iv

that full, or even approximately full, employmentis of rare and short-lived occurrence. Fluctuationsmay start briskly but seem to wear themselves outbefore they have proceeded to great extremes, and anintermediate situation which is neither desperate norsatisfactory is our normal lot. It is upon the fact thatfluctuations tend to wear themselves out before pro-ceeding to extremes and eventually to reverse them-selves, that the theory of business cycles having a regularphase has been founded. The same thing is true ofprices, which, in response to an initiating cause of dis-turbance, seem to be able to find a level at which theycan remain, for the time being, moderately stable.

Now, since these facts of experience do not followof logical necessity, one must suppose that the environ-ment and the psychological propensities of the modernworld must be of such a character as to produce theseresults. It is, therefore, useful to consider what hypo-thetical psychological propensities would lead to a stablesystem; and, then, whether these propensities can beplausibly ascribed, on our general knowledge of contem-porary human nature, to the world in which we live.

The conditions of stability which the foregoinganalysis suggests to us as capable of explaining theobserved results are the following:

(i) The marginal propensity to consume is suchthat, when the output of a given community increases(or decreases) because more (or less) employment isbeing applied to its capital equipment, the multiplierrelating the two is greater than unity but not very large.

(ii) When there is a change in the prospective yieldof capital or in the rate of interest, the schedule of themarginal efficiency of capital will be such that thechange in new investment will not be in great dispro-portion to the change in the former ; i.e. moderatechanges in the prospective yield of capital or in the rateof interest will not be associated with very great changesin the rate of investment.